USCS Determination for Soil w/ >50% Non Plastic Fines

[RobPE]

In ASTM D 4318, Test Method for Liquid Limit (LL), Plastic Limit (PL), and Plasticity Index (PI) of Soils, there is the option to state that the sample was tested and determined to be Non-Plastic, NP (section 19.1.1). This is done if either the LL or PL could not be determined or if the PL is equal or greater than the LL (ie: PI is negative).

The USCS method does not make mention of what to do when a soil is tested and determined to be NP but has a significant franction that passes a #200 (75 micron) sieve. For example we have had samples with greater than 50% passing a #200 sieve where the fines were determined to be NP. In some of these cases the LL could be determined but the PL could not (or the PL was greater than the LL).

How do you characterize this soil? If the LL is less than 50 and the PI is essentially zero, it would seem to be ML (silt). If the LL is greater than 50 but the PI can not be determined or is zero, would that be MH? What if the LL can not be determined?

 

[TDAA]

Typically, I have seen soils that are determined to be non-plastic (>50% -#200) classified as ML, with no reguard given to the LL. This is often the case because the result is typically reported as non-plastic, with no mention of the LL. This is reflected in ASTM D 4318, 19.1.1 Whenre if the LL or the PLcould not be determined, or the PL>=LL, the result is reported as Non-Plastic (ML).

 

[TDAA]

BTW, where are you working that you have soils that are "non-plastic" with LL>50. I'm sure it can happen, and probably happens everywhere I have not worked, but have not seen it personally.

 

[RobPE]

Thanks for the reply. I see what you're saying (we wind up calling these ML's too), but PI=0 is a valid result according to the plasticity chart, therefore if LL>50, the soil should classify as an MH which just doesn't make sense!@#$? I think this is a hole in the USCS classification system which ASTM should address at some point.

The samples we've tested recently, several of which have tested similar to what I described, are from a gold mine in Nevada (no gold in the samples).

Any other thoughts??

 

[dgillette]

Curious about gradation for these samples: %<5 micron and %<2 micron. Weathered volcanics?

Good point about the USCS. I think it and the plasticity chart would look different if we could go back and redesign them. The A-line is where it is because Casagrande drew it there one day (in the 1940s?), somewhat arbitrarily, before we had nearly as much experience with the mechanics of those soils. MH with LL=51 and PI=21 has a lot more in common with CH with LL=51 and PI=24 than it does with your samples, which may legitimately classify as MH. There are some extremely plastic montmorillonite-rich clays (determined by grain size and mineralogy) that classify as MH on the chart because they fall below the A-line. The cut between plastic SC and sandy CH is made at 50 percent fines, but with more than 35% clay fines, the sand pretty much floats in the matrix of fines (so the fines dominate the soil's behavior), so it might make more sense to make the cut there. We do Atterberg limits on the -#40 fraction (because Mr Atterberg the agronomist started out approximately that way), but nobody ever seems to know what the -#40 percentage is for their soils. My former boss once proposed changing that to the -#200 fraction, and got a whole lot of blank stares from the audience at the conference.

Oh, well, it's what we have to work with. Just shows the importance of not generalizing too much about behavior of broad groups, particularly MH or OH or OL.

 

[BigH]

This goes back to a thread that I started a while back on "whether a soil that is non-plastic can have a liquid limit." Opinion seemed to be equally divided - suggesting that the liquid limit can occur in a NP soil - I suggest that the liquid limit for a non-plastic soil is of little if any meaning. Good question - what would I do? I'd probably "cheat" and use M[sub]NP[/sub] - a non-standard term but perhaps one that should be created? ML has none to slight dry stregnth - MH has slight dry strength - so if one had to choose, I'd guess ML.

 

[GeoPaveTraffic]

If the LL is greater than 50 and the PL could not be dermined, I would go on and call it a MH. My reasoning is that a non-plastic silt with a LL that high should be looked at again and again to make sure that all the basis are covered; and I was always told to be very very careful of anything with a MH designation.

 

[TDAA]

It really doesn't matter if the chart has a "0" on it. I could draw one for you with a "-100". According to the text of the ASTM standard, if LL=PL, then it is reported as Non-Plastic.

As has been pointed out, you have to be careful in determining the behavior of the material, regardless of what the numbers tell you. Personally though, I think that even if you got the material to flow (has LL), if it is basically non-cohesive (unless moistures are at or above the LL) then it will act and classify as a ML. As BigH said, there is little meaning to the LL in a NP soil.

Actually, now that I think of it, I probably had soils that would get a LL and be NP, when I was out east. The thing of it is, you can add quite a bit of water to a sample, turn it to jello, and run a LL, but that doesn’t mean it will hold together to get a PI.

 

[RobPE]

From all of your responses, I think the basic conclusion is that this is a hole in the USCS system. BigH, I think your suggestion to classify this soil as a Mnp is the best idea proposed. Both the ML (low plasticity) and MH (high plasticity) designations are not completely correct since this soil is NON plastic. As has been stated by several of you, the LL in this situation is more or less meaningless.

Geopavetraffic you mentioned to be careful with soils with an MH designation. What were you referring to? Also, you mentioned that these soils should be looked at repeatedly to make sure that all of the bases are covered. With respect to this, we've tested several samples of this soil and have come up with the same results.

 

[dgillette]

Don't know what you plan to do with this material or what its geologic origin is, but

It may have a weird compaction moisture-density relationship and making compaction control more difficult,

It may be more compressible than is typical,

It is probably VERY easily eroded, and

Typical strengths may not apply, if for example, there is a lot of mica that could reduce the friction angle.

That very high LL means it can absorb/adsorb a lot of water somewhere before it can start acting "liquid." Could this make freeze-thaw problems worse?

This is strange enough material that you have piqued my curiosity, and probably that of others. Is it diatomaceous earth? Decomposed volcanics? Does it shrink and swell on drying and wetting? Is it residual soil, rather than lacustrine, alluvial, etc.? What part of what country?

DRG

 

[GeoPaveTraffic]

RobPE,

dgillette, pretty much covered the specific items that would be of concern. My point was "don't call this soil xxx (ML, MH, Non-plastic, whatever) and think about it just like all other soil of xxx type. This is an unusual material that, since you asked here; you don't see very often. Keep that in mind as you go through the design and particularly during construction.

Make sure that the personel in the field know about this unusuall material, that they are looking for it, and carefully monitoring how it responds to construction activities. You may find that it has some properties that cause problems or unexpected behavior during construction.

 

[jdonville]

RobPE,

From the description of the source (gold mine in Nevada) I am wondering if these are mine tailings?

As another question, would you be able to post (to add grist to the mill) the raw sieve and hydrometer data, along with the raw Atterberg data? Then we could all crunch the numbers and offer more interpretation on this unusual material.

Jeff

 

[RobPE]

Sorry to be so late in replying. Here's the sieve and hydrometer data for one of the samples we've been discussing:
mm % Passing
100 100.00
50.0 100.00
37.5 100.00
25.0 100.00
19.0 100.00
9.5 100.00
4.75 100.00
2.00 100.00
0.850 99.73
0.425 97.37
0.250 89.08
0.125 72.58
0.075 57.94
0.042899903 52.95
0.030722147 49.49
0.022214588 43.17
0.020036663 40.73
0.014478856 34.21
0.012039593 28.51
0.008728438 20.37
0.006246618 16.29
0.00312723 6.52
0.001344744 4.07

The other important piece of information is the liquid limit equal to 60% and no plastic limit.

The samples are from a waste dump in a gold mine in Nevada. This dump has been experiencing long term strain softening (time span over 10 years) that has led to recent major displacements. The samples in question are in fact weathered volcanics as you guessed dgillette. This material is at the base of the dump and as it turns out, is the weakest material in the dump profile. I'm not sure of the mineraloy of the samples, but I suspect that there is some montmorillonite and rock flour in these samples.

In addition to the sieves and PI's, we've also done some slow speed direct shear tests to determine the residual shear strength. Based on back analyses and the results of the testing, we've been able to come up with a model that reflects what has happened in the field. With this model, we've forward analyzed dump configurations that should be stable for the long term.

 

[BigH]

With a "LL" that high, you were actually able to keep the groove in the cup open when cutting it? Have you checked the LL vs a fall cone test? You indicate that you might have montmorillonite in the sample - it seems funny, then, that you have no PI. (p.s., I don't believe in reporting hydrometers or even sieves to the nearest 1/100th of a percentage - especially given the assumptions in a hydrometer). Again, with no PI, I think that LL is meaningless. The most important aspect you did was to do the slow direct shear tests to find out residual strength although the full softened strength values might be more appropriate (depends on the level of strain expected).

 

[TDAA]

So, based on the hydro, you have about 43% silt sized particles (AASHTO uses <0.075mm, >0.002mm), and overall it has >50% fines. Of course, USCS actually classifies based on behavior, not specifically on grain size. The behavior that you have indicated is one of silt (ML), and that is how the standard indicates this material should be classified (no PI = non-plastic = ML). I understand that people may not like the way some of the tests are run, or the way the charts are laid out, but this sample does not get lost in the system, if you actually read the standards. If you want to change the system, start talking with the ASTM comity that oversees the method.

I am not sure why this sample should see additional scrutiny beyond the fact that ML soils are difficult to construct with and are generally weak materials. Are the properties different than the other classifications? Sure. Of course it doesn’t take long to figure that out when you look at the material or watch a contractor try to compact this “fine grained material” with a sheep’s foot roller.

 

[RobPE]

A few responses: our lab techs report that the LL was successfully measured at 60% - haven't checked the LL vs falling cone test but will look into that - I'm not sure of the mineralogy of the samples (they may or may not have montmorrilonite) - I also don't believe in reporting ANYthing to the nearest thousandth (I just pasted values from an excel file). We've also looked at Mesri's approach of determining residual shear strength based on PI. Our residual test values approach Mesri's fully softened values, which are appropriate considering the long term strain history of these dumps.

I'm always impressed by the insight of those of you that regularly submit posts on this website.

 

[RobPE]

One more response to TDAA: If you strictly follow the ASTM flowchart for determining the USCS classification for this soil, it does not lead to an ML designation. ML is a very reasonable classification for this soil, but if you have good lab techs who follow the ASTM procedures verbatim, then this soil gets lost in the classification.

 

[BigH]

Why don't we ask ASTM? I know they "ALWAYS" like questions - like the one I asked when they wanted me to report concrete strengths to the nearest 79 kPa!! (Thanks that they have now changed it to the nearest 100 kPa (0.1 MPa).

 

[TDAA]

RobPE:

If you follow through ASTM D 4318, specifically 19.1.1, you would not have a reported value for a LL. This section indicates to "report the soil as nonplastic, NP" when the either the LL or PL cannot be determined or the PL>=LL. The same think is reiterated in section 20.

So, I run the tests in my lab for you. 60% passing the #200, LL=52, PI could not be determined. According to the standard, I report to you that the soil has 60% passing the #200 sieve, and is NP. How do you classify this?

My point is that the concentration appears to be on the flow chart of how to classify, and that people are trying to use data that "is not there". That is, it is not reported.

Don’t get me wrong, I believe in using ALL the information available to help determine the engineering properties, including information that is not always reported; however, I do not feel that this soil gets lost in the classification system. It is a matter of interpreting the index-testing standard correctly.